Changing the subject a little, I thought that turbine blade root cross sectional areas were sized so as to fail in tension in the event of an overspeed. Thus hopefully limiting the extent of overspeed
Sounds like a great idea (fusing) but it opens up other cans of worms.
A common blade attachment max stress level at 100% operating RPM in tension gives a 50% margin in overspeed. Tensile margins are nice when one considers their interaction with vibratory stress levels in a spinning machine (Goodman Diagram) but if you were to design every disk with something greater than the blade margin you would have an extremely heavy engine.
Then one has to consider the mass of the blade being released in a fused overspeed, and whether its practical to design containment for that event. In the end many things are considered together in protecting against overspeeds with the majority of the consideration being placed on limiting the speed the machine can get without being uncontained.
By-the-way, it's more likely that the stretch in the disk will open up the fir tree and release the whole blade before the attachment would fail (common in a disk over heat event)